Dynamic self-adaptive suppression method and device for alarm of Internet of Things

Abstract

The invention relates to an Internet of Things alarm dynamic self-adaptive suppression method and device, and belongs to the technical field of self-adaptive suppression identification, and the method comprises the steps that an Internet of Things cloud platform receives an alarm generated by an Internet of Things equipment cluster; extracting feature data of the alarm in real time, and querying a historical behavior mode of the corresponding Internet of Things equipment and the corresponding alarm; receiving the feature data and the historical behavior pattern, inputting the feature data and the historical behavior pattern into a pre-trained silent period prediction model to output a dynamic silent period suggestion value of the corresponding alarm, and checking whether the corresponding Internet of Things equipment and the corresponding alarm type are in a silent state; if the check result is that the device is not in the silent state, setting a silent period suppression corresponding alarm with the length being a dynamic silent period suggestion value, and reporting the alarm to operation and maintenance personnel; judging whether the silent period expires or not, and if the judgment result is that the silent period expires; and judging whether the fault of the corresponding alarm Internet of Things equipment is recovered, and if the judgment result is that the fault of the equipment is recovered, ending suppression.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of dynamic self-adaptive suppression, and particularly relates to an Internet of Things alarm dynamic self-adaptive suppression method and device. BACKGROUND

[0002] With the popularization of the Internet of Things technology, the number of terminal devices accessing the platform increases exponentially. When the device detects an abnormal state (such as excessively high temperature or excessive vibration), the device will send an alarm message to the Internet of Things platform. However, in actual application, the following technical pain points exist:

[0003] Repeated alarms are rampant: since the abnormal state of the device often lasts for a period of time, the device will periodically and continuously report alarms, resulting in a large number of repeated contents received by the platform in a short time, causing an "alarm storm".

[0004] Static rules are not flexible: existing solutions mostly use a static silence period mechanism, that is, the same type of alarms are suppressed within a fixed time, such as 5 minutes. This method lacks adaptability: if the silence period is too long, the follow-up confirmation information of important alarms will be lost, delaying the response; if the silence period is too short, the storm cannot be effectively suppressed, and the platform resources are still occupied.

[0005] Resource consumption is serious: the alarm storm will consume a large amount of network bandwidth, platform computing resources and storage resources, and may drown the truly important new alarms, causing "alarm fatigue" of the operation and maintenance personnel, and reducing the operation and maintenance efficiency.

[0006] Existing technologies usually provide an edge computing pre-filtering scheme, but the filtering rules are still pre-configured statically, and cannot be adaptively adjusted according to the global situation and device individuality. Or, a threshold is set for alarm judgment, but the threshold is also static. Therefore, there is an urgent need for a solution that can intelligently perceive the alarm context and dynamically adjust the suppression strategy. SUMMARY

[0007] In view of the deficiencies of the above prior art, the purpose of the application is to provide an Internet of Things alarm dynamic self-adaptive suppression method and device, which realizes real-time self-adaptive adjustment of the alarm suppression strategy through multi-dimensional feature fusion, dynamic model decision and feedback optimization, and solves the pain point that the static rules in the prior art are rigid and cannot cope with complex scenarios.

[0008] In a first aspect of the application, an Internet of Things alarm dynamic self-adaptive suppression method is provided, which is applied to a system composed of an Internet of Things device cluster and an Internet of Things cloud platform, and the method comprises:

[0009] Step 1: The Internet of Things cloud platform receives alarms generated by the Internet of Things device cluster;

[0010] Step two: extract the feature data of the alarm in real time, query the historical behavior mode of the corresponding Internet of Things device and the corresponding alarm from the historical database;

[0011] Step three: receive the feature data and the historical behavior mode, input the feature data and the historical behavior mode into the pre-trained silent period prediction model to output the dynamic silent period suggestion value of the corresponding alarm, and check whether the corresponding Internet of Things device and the corresponding alarm type are in a silent state;

[0012] Step four: if the check result is that the corresponding Internet of Things device and the corresponding alarm type are not in a silent state, set a silent period with a length of the dynamic silent period suggestion value to suppress the corresponding alarm, and report the alarm to the operation and maintenance personnel;

[0013] Step five: determine whether the silent period is expired, and if the determination result is that the silent period is expired;

[0014] Step six: determine whether the corresponding alarm Internet of Things device fault is recovered, and if the determination result is that the corresponding alarm Internet of Things device fault has been recovered, end the suppression.

[0015] Further, the above-mentioned Internet of Things alarm dynamic adaptive suppression method further comprises:

[0016] determine whether the silent period is expired, and if the determination result is that the silent period is expired;

[0017] determine whether the corresponding alarm Internet of Things device fault is recovered, and if the determination result is that the corresponding alarm Internet of Things device fault has not been recovered;

[0018] repeat steps one to six until the corresponding Internet of Things device fault is recovered;

[0019] Among them, the feature data includes: Internet of Things device ID, alarm type, timestamp, current alarm instantaneous frequency and system CPU / memory load; the historical behavior mode includes: alarm frequency, alarm period and alarm duration.

[0020] Further, the above-mentioned Internet of Things alarm dynamic adaptive suppression method further comprises:

[0021] check whether the corresponding Internet of Things device and the corresponding alarm type are in a silent state;

[0022] if the check result is that the corresponding Internet of Things device and the corresponding alarm type are in a silent state, discard or aggregate the corresponding alarm.

[0023] Further, the above-mentioned Internet of Things alarm dynamic adaptive suppression method further comprises:

[0024] continuously monitor the effect of the suppression strategy to generate feedback data;

[0025] The feedback learning engine uses a reinforcement learning algorithm to perform online optimization and adjustment on the pre-trained quiet period prediction model parameters based on the feedback data.

[0026] The feedback data at least includes: actual suppression alarm number and whether important alarms are mistakenly suppressed.

[0027] Further, in the above-mentioned alarm dynamic adaptive suppression method of the Internet of Things, the real-time extraction of the feature data of the alarm and the query of the historical behavior mode corresponding to the Internet of Things device and the corresponding alarm from the historical database are realized through a behavior analysis engine in the Internet of Things cloud platform.

[0028] Further, in the above-mentioned alarm dynamic adaptive suppression method of the Internet of Things, the receiving of the feature data and the historical behavior mode, the input of the feature data and the historical behavior mode into a pre-trained quiet period prediction model to output a dynamic quiet period suggestion value corresponding to the alarm, and the checking of whether the corresponding Internet of Things device and the corresponding alarm type are in a quiet state are realized through a dynamic decision engine in the Internet of Things cloud platform.

[0029] Further, in the above-mentioned alarm dynamic adaptive suppression method of the Internet of Things, the training step of the pre-trained quiet period prediction model comprises:

[0030] Collecting historical alarm data and associated system load indicators;

[0031] Extracting instantaneous alarm frequency and historical behavior mode as feature vectors, and labeling optimal quiet period labels;

[0032] Taking the feature vectors as input and the optimal quiet period as output, a supervised learning or reinforcement learning model is trained to obtain a pre-trained quiet period prediction model.

[0033] The second aspect of the application also proposes an alarm dynamic adaptive suppression device of the Internet of Things, which is applied to a system composed of an Internet of Things device cluster and an Internet of Things cloud platform, and the device comprises:

[0034] The receiving module is used for receiving alarms generated by the Internet of Things device cluster by the Internet of Things cloud platform.

[0035] The extraction module is used for real-time extraction of the feature data of the alarm and query of the historical behavior mode corresponding to the Internet of Things device and the corresponding alarm from the historical database.

[0036] The output module is used for receiving the feature data and the historical behavior mode, inputting the feature data and the historical behavior mode into a pre-trained quiet period prediction model to output a dynamic quiet period suggestion value corresponding to the alarm, and checking whether the corresponding Internet of Things device and the corresponding alarm type are in a quiet state.

[0037] The setting module is configured to set a silence period with a length of a dynamic silence period suggestion value to suppress the corresponding alarm if the check result is that the corresponding Internet of Things device and the corresponding alarm type are not in a silence state, and report the alarm to an operation and maintenance personnel;

[0038] The judgment module is configured to judge whether the silence period is expired or not, and if the judgment result is that the silence period is expired.

[0039] The ending suppression module is configured to judge whether the corresponding alarm Internet of Things device failure is recovered or not, and if the judgment result is that the corresponding alarm Internet of Things device failure has been recovered, the suppression is ended.

[0040] The third aspect of the present application further provides an electronic device, comprising a processor and a memory.

[0041] The processor is configured to execute any one of the above Internet of Things alarm dynamic adaptive suppression methods by calling programs or instructions stored in the memory.

[0042] The fourth aspect of the present application further provides a computer readable storage medium, which stores programs or instructions, and the programs or instructions make the computer execute any one of the above Internet of Things alarm dynamic adaptive suppression methods.

[0043] The present application has the following beneficial effects:

[0044] 1) Dynamic adaptability: change the fixed silence period mode, dynamically calculate and adjust the optimal suppression strategy according to the real-time alarm frequency, device historical behavior mode and system current load.

[0045] 2) Intelligent suppression: establish a behavior baseline for each device and each alarm type through a machine learning algorithm to realize individualized and accurate alarm suppression.

[0046] 3) Resource optimization: significantly reduce the amount of redundant alarm data processed by the platform, reduce the consumption of network, computing and storage resources, and improve the overall stability of the system.

[0047] 4) Operation and maintenance efficiency: provide more concise and important alarm information flow for operation and maintenance personnel, reduce interference, and improve the efficiency of fault diagnosis and processing. BRIEF DESCRIPTION OF DRAWINGS

[0048] The accompanying drawings are included to provide a further understanding of the present application, and are incorporated herein and constitute a part of the detailed description. It should be apparent to those skilled in the art that the accompanying drawings are only some embodiments of the present application described in the following description. Those skilled in the art can also obtain the remaining drawings from the accompanying drawings.

[0049] Figure 1A method diagram for dynamically adaptive suppression of Internet of Things alarms provided for an embodiment of the present application is provided;

[0050] Figure 2 A corresponding alarm Internet of Things device failure has not recovered suppression method diagram provided for an embodiment of the present application is provided;

[0051] Figure 3 A corresponding Internet of Things device and corresponding alarm type has been in a silent state suppression method diagram provided for an embodiment of the present application is provided;

[0052] Figure 4 A method diagram for optimizing and adjusting the parameters of a pre-trained silent period prediction model provided for an embodiment of the present application is provided;

[0053] Figure 5 A training method diagram for a pre-trained silent period prediction model provided for an embodiment of the present application is provided;

[0054] Figure 6 A device diagram for dynamically adaptive suppression of Internet of Things alarms provided for an embodiment of the present application is provided;

[0055] Figure 7 A schematic block diagram of an electronic device provided for an embodiment of the present application is provided. DETAILED DESCRIPTION

[0056] In order to make the personnel in the art better understand the technical solutions in the embodiments of the present application, the technical solutions of the present application will be described clearly and completely below in conjunction with the drawings. Obviously, the described embodiments are part of the embodiments of the present application, rather than all the embodiments. It should be understood that these descriptions are only exemplary and are not intended to limit the scope of the present application. Based on the embodiments of the present application, all the remaining embodiments obtained by those of ordinary skill in the art without making creative efforts should fall within the scope of protection of the present application.

[0057] In addition, in the following description, the description of well-known structures and technologies is omitted to avoid unnecessary confusion of the concepts disclosed in the present application.

[0058] In the description of the present application, the terms "first", "second", "third" are only for descriptive purposes, and cannot be understood as indicating or implying relative importance. The terms "mounting", "connecting", "connecting" should be broadly understood, for example, it can be fixedly connected, or it can be detachably connected, or integrally connected; it can be mechanically connected, or it can be electrically connected; it can be directly connected, or it can be indirectly connected through an intermediate medium, or it can be the communication between two elements. For those of ordinary skill in the art, the specific meaning of the above terms in the present application can be understood according to the specific circumstances.

[0059] The exemplary embodiments will be described in detail herein with reference to the attached drawings. In the following description, like reference numerals refer to like elements throughout the description. The following exemplary embodiments are described with reference to the drawings. The embodiments described in these exemplary embodiments do not represent all of the embodiments consistent with the present application. Rather, they are merely examples consistent with some aspects of the present application as detailed in the appended claims.

[0060] The present application provides an Internet of Things alarm dynamic adaptive suppression method, device, electronic equipment and storage medium, through multi-dimensional feature fusion, dynamic model decision and feedback optimization, realizing real-time adaptive adjustment of alarm suppression strategy, solving the pain point that the static rule in the prior art is rigid and cannot cope with complex scenes.

[0061] Before introducing the embodiments of the present application, first introduce the professional terms related to the present application.

[0062] Alarm storm: refers to the phenomenon that the Internet of Things terminal device continuously triggers and reports a large number of alarm messages with the same or similar content in a short time, consuming network and platform resources.

[0063] Static silence period: a pre-set, fixed-length alarm suppression time window in the prior art. In the window, similar alarms will be ignored or suppressed by the platform.

[0064] Behavior baseline: a data model obtained by learning from historical data, used to describe the behavior mode (such as alarm triggering frequency, period, duration, etc.) of a single Internet of Things device in a normal state or a specific alarm in a typical scenario.

[0065] Dynamic silence period: refers to the optimal alarm suppression time window dynamically calculated and generated by the algorithm according to the real-time context (such as current alarm frequency, system load, historical behavior), and the length is not a fixed value.

[0066] Reinforcement learning: a machine learning paradigm, in which an agent learns to take the optimal strategy in a specific environment by interacting with the environment and obtaining rewards (positive or negative).

[0067] Method embodiment

[0068] The first aspect of the present application provides an Internet of Things alarm dynamic adaptive suppression method, applied to a system composed of an Internet of Things device cluster and an Internet of Things cloud platform, combining Figure 1 The method comprises the following steps:

[0069] S1: The Internet of Things cloud platform receives alarms generated by the Internet of Things device cluster.

[0070] Specifically, in the embodiment of the present application, the Internet of Things cloud platform receives original alarm streams generated by all Internet of Things devices in the Internet of Things device cluster.

[0071] For example, the smart meter DT-001 starts to report an overvoltage alarm to the Internet of Things cloud platform once every second due to a line fault, and the Internet of Things cloud platform receives the overvoltage alarm reported by the smart meter DT-001 once.

[0072] S2: Real-time extraction of feature data of the alarm, and query of historical behavior patterns of the corresponding Internet of Things device and the corresponding alarm from a historical database.

[0073] Specifically, in the embodiment of the present application, the behavior analysis engine in the Internet of Things cloud platform extracts feature data of the alarm in real time, and queries historical behavior patterns of the corresponding Internet of Things device and the corresponding alarm from a historical database, wherein the feature data includes: an Internet of Things device ID, an alarm type, a timestamp, a current alarm instantaneous frequency, and system CPU / memory load; and the historical behavior patterns include: an alarm frequency, an alarm period, and an alarm duration.

[0074] For example, the behavior analysis engine in the Internet of Things cloud platform receives the overvoltage alarm reported by the smart meter DT-001, extracts feature data: device ID = DT-001, alarm type = OverVoltage, timestamp = t0, system CPU / memory load, such as CPU usage rate of 65%, calculates the instantaneous frequency of the alarm, such as the number of times of this type of alarm in the past 10 seconds, instantaneous frequency f = 1, queries the historical behavior patterns of the corresponding alarm of the corresponding Internet of Things device DT-001 from a historical database, finds that the historical average frequency of the alarm is 0.01 times / minute, and determines that the behavior baseline belongs to a rare alarm.

[0075] S3: Receiving the feature data and the historical behavior patterns, inputting the feature data and the historical behavior patterns into a pre-trained silent period prediction model to output a dynamic silent period suggestion value of the corresponding alarm, and checking whether the corresponding Internet of Things device and the corresponding alarm type are in a silent state.

[0076] Specifically, in the embodiment of the present application, the behavior analysis engine in the Internet of Things cloud platform inputs the feature data and the historical behavior patterns into a dynamic decision engine in the Internet of Things cloud platform, the dynamic decision engine in the Internet of Things cloud platform receives the feature data and the historical behavior patterns, inputs the feature data and the historical behavior patterns into a pre-trained silent period prediction model to output a dynamic silent period suggestion value of the corresponding alarm, and the dynamic decision engine in the Internet of Things cloud platform checks whether the corresponding Internet of Things device and the corresponding alarm type are in a silent state.

[0077] For example, the behavior analysis engine in the IoT cloud platform inputs a rare alarm corresponding to an instantaneous frequency f=1, CPU utilization of 65%, and historical average alarm frequency of 0.01 times / minute into a pre-trained silent period prediction model. The pre-trained silent period prediction model judges that this trigger may be the beginning of a major event, so it outputs a short dynamic silent period T_silence=30S to avoid missing any subsequent important changes. The dynamic decision engine in the IoT cloud platform checks whether the corresponding IoT device and the corresponding alarm type are already in a silent state.

[0078] S4: If the inspection result shows that the corresponding IoT device and the corresponding alarm type are not in a silent state, set a silent period with a length equal to the recommended dynamic silent period to suppress the corresponding alarm and report the alarm to the operation and maintenance personnel.

[0079] Specifically, in this embodiment of the invention, if the dynamic decision engine in the IoT cloud platform determines that the corresponding IoT device and the corresponding alarm type are not in a silent state, a silent period of length equal to the recommended value of the dynamic silent period is set to suppress the corresponding alarm, and the alarm is reported to the operation and maintenance personnel.

[0080] For example, the dynamic decision engine in the IoT cloud platform checks that the DT-001 OverVoltage alarm currently has no silent period and is not in a silent state. It sets a silent period with a length of dynamic silent period T_silence = 30S to suppress the corresponding alarm, with an expiration time of t0+30S, and reports the alarm to the operation and maintenance personnel.

[0081] S5: Determine if the silent period has expired. If the result is that the silent period has expired.

[0082] Specifically, in this embodiment of the invention, the dynamic decision engine in the IoT cloud platform determines whether the quiet period has expired. If the determination result is that the quiet period has expired...

[0083] For example, the dynamic decision engine in the IoT cloud platform determines whether t0+30S has been reached. If the determination result is that t0+30S has been reached...

[0084] S6: Determine whether the corresponding alarm IoT device fault has been recovered. If the result is that the corresponding alarm IoT device fault has been recovered, then end the suppression.

[0085] Specifically, in this embodiment of the invention, the dynamic decision engine in the IoT cloud platform continues to determine whether the corresponding alarm IoT device fault has been recovered. If the determination result is that the corresponding alarm IoT device fault has been recovered, the suppression ends.

[0086] For example, if the dynamic decision engine in the IoT cloud platform continues to determine that the overvoltage alarm of the smart meter DT-001 has been restored, then the suppression will end.

[0087] Furthermore, the aforementioned IoT alarm dynamic adaptive suppression method, combined with Figure 2 It also includes:

[0088] S21: Determine whether the silent period has expired. If the result is that the silent period has expired.

[0089] S22: Determine whether the corresponding alarm IoT device fault has been recovered. If the result is that the corresponding alarm IoT device fault has not been recovered.

[0090] Specifically, in this embodiment of the invention, the dynamic decision engine in the IoT cloud platform continues to determine whether the corresponding alarm IoT device fault has been recovered. If the determination result is that the corresponding alarm IoT device fault has not been recovered.

[0091] S23: Repeat steps S1 to S6 until the corresponding IoT device fault is resolved.

[0092] Specifically, in this embodiment of the invention, the dynamic decision engine in the IoT cloud platform determines whether the quiet period has expired. If the result is that the quiet period has expired, the dynamic decision engine in the IoT cloud platform continues to determine whether the corresponding alarm IoT device fault has been recovered. If the result is that the corresponding alarm IoT device fault has not been recovered, steps S1 to S6 are repeated until the corresponding IoT device fault is recovered.

[0093] For example, the dynamic decision engine in the IoT cloud platform determines whether t0+30S has been reached. If the result is that t0+30S has been reached, the dynamic decision engine in the IoT cloud platform continues to determine that the overvoltage alarm of smart meter DT-001 has not yet recovered, and the second OverVoltage alarm has arrived. The behavior analysis engine calculates that the instantaneous frequency f has risen to 10 times / 10 seconds, the CPU utilization rate is 68%, and the rare alarm corresponding to the historical average alarm frequency of 0.01 times / minute is input into the pre-trained silent period prediction model. The pre-trained silent period prediction model senses... The alarm frequency spiked dramatically, far exceeding the baseline behavior patterns observed in the past, indicating an alarm storm had formed. Simultaneously, CPU utilization also increased slightly. To mitigate the impact of this storm on the IoT cloud platform, the pre-trained silent period prediction model outputs a longer dynamic silent period T_silence = 300 seconds. If the meter malfunction persists after 300 seconds, the first alarm after that time will be re-evaluated by the pre-trained silent period prediction model. This model may output a new silent period, such as 10 minutes, and repeat the suppression process. Simultaneously, the feedback learning engine evaluates the suppression of 300 alarms over the past 300 seconds, confirming its effective protection of the IoT cloud platform. This provides a positive reward to the pre-trained silent period prediction model, reinforcing this decision-making process.

[0094] Furthermore, the aforementioned IoT alarm dynamic adaptive suppression method, combined with Figure 3 It also includes two steps, S31 and S32:

[0095] S31: Check whether the corresponding IoT device and the corresponding alarm type are in a silent state;

[0096] S32: If the check result shows that the corresponding IoT device and the corresponding alarm type are in a silent state, then discard or aggregate the corresponding alarm.

[0097] Specifically, in this embodiment of the invention, the dynamic decision engine in the IoT cloud platform checks whether the corresponding IoT device and the corresponding alarm type are in a silent state. If the dynamic decision engine in the IoT cloud platform determines that the corresponding IoT device and the corresponding alarm type are in a silent state, then the corresponding alarm is discarded or aggregated.

[0098] Figure 4 This diagram illustrates a method for optimizing and adjusting the parameters of a pre-trained silent period prediction model, as provided in an embodiment of the present invention.

[0099] Furthermore, the aforementioned IoT alarm dynamic adaptive suppression method, combined with Figure 4 It also includes two steps, S41 and S42:

[0100] S41: Continuously monitor the effectiveness of the suppression strategy and generate feedback data;

[0101] S42: The feedback learning engine uses reinforcement learning algorithms to optimize and adjust the parameters of the pre-trained silent period prediction model online based on feedback data;

[0102] The feedback data should include at least the actual number of alarms suppressed and whether any important alarms were mistakenly suppressed.

[0103] Specifically, in this embodiment of the invention, the feedback learning engine in the IoT cloud platform continuously monitors the effect of the suppression strategy and generates feedback data, such as the actual number of suppressed alarms and whether important alarms were mistakenly suppressed. Based on this effect data, the feedback learning engine uses reinforcement learning algorithms to optimize and adjust the parameters of the silent period prediction model online, forming a closed loop, making the prediction increasingly accurate.

[0104] Furthermore, in the aforementioned IoT alarm dynamic adaptive suppression method, the real-time extraction of alarm feature data and the querying of historical behavior patterns of corresponding IoT devices and alarms from the historical database are achieved through the behavior analysis engine in the IoT cloud platform.

[0105] Specifically, in this embodiment of the invention, the behavior analysis engine in the IoT cloud platform extracts the feature data of the alarm in real time and queries the historical behavior patterns of the corresponding IoT device and the corresponding alarm from the historical database. The feature data includes: IoT device ID, alarm type, timestamp, current alarm instantaneous frequency and system CPU / memory load; the historical behavior patterns include: alarm frequency, alarm cycle and alarm duration.

[0106] Furthermore, in the aforementioned IoT alarm dynamic adaptive suppression method, the feature data and historical behavior patterns are received, and the feature data and historical behavior patterns are input into a pre-trained silent period prediction model to output a dynamic silent period suggestion value for the corresponding alarm. Checking whether the corresponding IoT device and the corresponding alarm type are in a silent state is achieved through a dynamic decision engine in the IoT cloud platform.

[0107] Specifically, in this embodiment of the invention, the behavior analysis engine in the IoT cloud platform inputs feature data and historical behavior patterns into the dynamic decision engine in the IoT cloud platform. The dynamic decision engine in the IoT cloud platform receives the feature data and historical behavior patterns, inputs the feature data and historical behavior patterns into a pre-trained silent period prediction model, and outputs a dynamic silent period suggestion value for the corresponding alarm. The dynamic decision engine in the IoT cloud platform checks whether the corresponding IoT device and the corresponding alarm type are in a silent state.

[0108] Furthermore, in the aforementioned dynamic adaptive suppression method for IoT alarms, the training method of the pre-trained silent period prediction model is combined with... Figure 5 It also includes five steps, S51 to S53:

[0109] S51: Collect historical alarm data and related system load metrics;

[0110] S52: Extract instantaneous alarm frequency and historical behavior patterns as feature vectors, and label the optimal silent period.

[0111] S53: Using feature vectors as input and optimal quiet period as output, train a supervised learning or reinforcement learning model to obtain a pre-trained quiet period prediction model.

[0112] Specifically, in this embodiment of the invention, a pre-trained silent period prediction model is obtained through the above steps. The pre-trained silent period prediction model is based on the frequency and load of real-time perception. The pre-trained silent period prediction model calculates an optimal and personalized dynamic silent period for each alarm in real time, transforming the suppression strategy from manually configured and static to algorithm-generated and dynamic. This achieves millisecond-level adaptive adjustment of the strategy. Moreover, the dynamic decision is not based on a single rule, but depends on the input of two dimensions: instantaneous frequency, system load, and historical behavior, resulting in more accurate decision-making.

[0113] Device Examples

[0114] Figure 6 This is a diagram of an IoT alarm dynamic adaptive suppression device provided in an embodiment of the present invention.

[0115] A second aspect of the present invention also proposes an IoT alarm dynamic adaptive suppression device, applied to a system composed of an IoT device cluster and an IoT cloud platform, the device comprising:

[0116] Receiver module 61: Used for receiving alarms generated by the IoT device cluster on the IoT cloud platform.

[0117] Specifically, in this embodiment of the invention, the receiving module 61 in the IoT cloud platform receives the raw alarm streams generated by all IoT devices in the IoT device cluster.

[0118] Extraction module 62: Used to extract feature data of alarms in real time and query the historical behavior patterns of corresponding IoT devices and alarms from the historical database.

[0119] Specifically, in this embodiment of the invention, the extraction module 62 in the behavior analysis engine of the IoT cloud platform extracts the feature data of the alarm in real time and queries the historical behavior patterns of the corresponding IoT device and the corresponding alarm from the historical database. The feature data includes: IoT device ID, alarm type, timestamp, current alarm instantaneous frequency and system CPU / memory load; the historical behavior patterns include: alarm frequency, alarm cycle and alarm duration.

[0120] Output module 63: Used to receive feature data and historical behavior patterns, input the feature data and historical behavior patterns into the pre-trained silent period prediction model, output the dynamic silent period suggestion value of the corresponding alarm, and check whether the corresponding IoT device and the corresponding alarm type are in a silent state.

[0121] Specifically, in this embodiment of the invention, the behavior analysis engine in the IoT cloud platform inputs feature data and historical behavior patterns into the dynamic decision engine in the IoT cloud platform. The dynamic decision engine in the IoT cloud platform receives the feature data and historical behavior patterns, inputs the feature data and historical behavior patterns into the pre-trained silent period prediction model, and the output module 63 outputs the dynamic silent period suggestion value corresponding to the alarm. The dynamic decision engine in the IoT cloud platform checks whether the corresponding IoT device and the corresponding alarm type are in a silent state.

[0122] Setting module 64: If the inspection result shows that the corresponding IoT device and the corresponding alarm type are not in a silent state, it sets a silent period with a length of the recommended dynamic silent period to suppress the corresponding alarm and reports the alarm to the operation and maintenance personnel.

[0123] Specifically, in this embodiment of the invention, if the dynamic decision engine in the IoT cloud platform determines that the corresponding IoT device and the corresponding alarm type are not in a silent state, the setting module 64 sets a silent period with a length equal to the recommended value of the dynamic silent period to suppress the corresponding alarm and reports the alarm to the operation and maintenance personnel.

[0124] Judgment Module 65: Used to determine whether the silent period has expired. If the judgment result is that the silent period has expired.

[0125] Specifically, in this embodiment of the invention, the judgment module in the dynamic decision engine of the IoT cloud platform determines whether the quiet period has expired. If the judgment result is that the quiet period has expired, the following steps are continued.

[0126] End suppression module 66: Used to determine whether the corresponding alarm IoT device fault has been recovered. If the determination result is that the corresponding alarm IoT device fault has been recovered, then the suppression ends.

[0127] Specifically, in this embodiment of the invention, the dynamic decision engine in the IoT cloud platform continues to determine whether the corresponding alarm IoT device fault has been recovered. If the determination result is that the corresponding alarm IoT device fault has been recovered, then the suppression module 66 ends the suppression.

[0128] A third aspect of the present invention also provides an electronic device comprising: a processor and a memory;

[0129] The processor executes one of the above-mentioned IoT alarm dynamic adaptive suppression methods by calling programs or instructions stored in memory.

[0130] In a fourth aspect, the present invention also provides a computer-readable storage medium storing a program or instructions that cause a computer to execute any of the above-mentioned IoT alarm dynamic adaptive suppression methods.

[0131] Figure 7 This is a schematic block diagram of an electronic device provided in an embodiment of the present invention.

[0132] like Figure 7 As shown, the electronic device includes at least one processor 701, at least one memory 702, and at least one communication interface 703. The various components of the electronic device are coupled together via a bus system 704. The communication interface 703 is used for information transmission with external devices. It is understood that the bus system 704 is used to implement communication between these components. In addition to a data bus, the bus system 704 also includes a power bus, a control bus, and a status signal bus. However, for clarity, ... Figure 7 The general designated all buses as Bus System 704.

[0133] It is understood that the memory 702 in this embodiment may be volatile memory or non-volatile memory, or may include both volatile and non-volatile memory.

[0134] In some implementations, memory 702 stores elements such as executable units or data structures, or subsets thereof, or extended sets thereof: operating systems and applications.

[0135] The operating system includes various system programs, such as the framework layer, core library layer, and driver layer, used to implement various basic business functions and handle hardware-based tasks. The application programs include various applications, such as media players and browsers, used to implement various application functions. The program implementing any method in the IoT alarm dynamic adaptive suppression method provided in this embodiment of the invention can be included in the application programs.

[0136] In this embodiment of the invention, the processor 701 executes the steps of various embodiments of the IoT alarm dynamic adaptive suppression method provided in this embodiment of the invention by calling the program or instructions stored in the memory 702, specifically, the program or instructions stored in the application program.

[0137] Step 1: The IoT cloud platform receives alarms generated by the IoT device cluster;

[0138] Step 2: The behavior analysis engine in the IoT cloud platform extracts the feature data of the alarm in real time and queries the historical behavior patterns of the corresponding IoT devices and alarms from the historical database;

[0139] Step 3: Receive feature data and historical behavior patterns, input the feature data and historical behavior patterns into the pre-trained silent period prediction model, output the dynamic silent period suggestion value of the corresponding alarm, and check whether the corresponding IoT device and the corresponding alarm type are in a silent state.

[0140] Step 4: If the inspection results show that the corresponding IoT device and the corresponding alarm type are not in a silent state, set a silent period with a length equal to the recommended dynamic silent period to suppress the corresponding alarm, and report the alarm to the operation and maintenance personnel.

[0141] Step 5: Determine if the quiet period has expired. If the result is that the quiet period has expired;

[0142] Step 6: Determine whether the corresponding alarm IoT device fault has been resolved. If the determination result is that the corresponding alarm IoT device fault has been resolved, then end the suppression.

[0143] Any method in the IoT alarm dynamic adaptive suppression method provided in this embodiment of the invention can be applied to, or implemented by, the processor 701. The processor 701 can be an integrated circuit chip with signal processing capabilities. During implementation, each step of the above method can be completed by the integrated logic circuitry in the hardware of the processor 701 or by instructions in software form. The processor 701 can be a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. The general-purpose processor can be a microprocessor or any conventional processor.

[0144] The steps of any method in the IoT alarm dynamic adaptive suppression method provided in this embodiment of the invention can be directly implemented by a hardware decoding processor, or implemented by a combination of hardware and software units in the decoding processor. The software units can reside in random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, registers, or other mature storage media in the art. This storage medium is located in memory 702, and processor 701 reads the information in memory 702 and combines it with its hardware to complete the steps of the method.

[0145] Those skilled in the art will understand that although some embodiments described herein include certain features included in other embodiments but not others, combinations of features from different embodiments are meant to be within the scope of the invention and form different embodiments.

[0146] Those skilled in the art will understand that the descriptions of the various embodiments have different focuses, and for parts not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0147] Although embodiments of the present invention have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of the invention. All such modifications and variations fall within the scope defined by the appended claims. The above are merely specific embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in the present invention, and these modifications or substitutions should all be covered within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

[0148] The above are merely specific embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in the present invention, and these modifications or substitutions should all be covered within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for dynamically adaptive suppression of alarms in an Internet of Things, characterized in that, The method is applied to a system composed of an Internet of Things device cluster and an Internet of Things cloud platform, and the method comprises the following steps: Step 1: The Internet of Things cloud platform receives an alarm generated by the Internet of Things device cluster; Step 2: Real-time feature data of the alarm is extracted, and a historical behavior mode of the corresponding Internet of Things device and the corresponding alarm is queried from a historical database; Step 3: The feature data and the historical behavior mode are received, the feature data and the historical behavior mode are input into a pre-trained silence period prediction model to output a dynamic silence period suggestion value of the corresponding alarm, and it is checked whether the corresponding Internet of Things device and the corresponding alarm type are in a silence state; Step 4: If the checking result is that the corresponding Internet of Things device and the corresponding alarm type are not in the silence state, a silence period with a length of the dynamic silence period suggestion value is set to suppress the corresponding alarm, and the alarm is reported to an operation and maintenance personnel; Step 5: Whether the silence period is expired is judged, and if the judgment result is that the silence period is expired; Step 6: Whether a fault of the corresponding alarm Internet of Things device is recovered is judged, and if the judgment result is that the fault of the corresponding alarm Internet of Things device is recovered, the suppression is ended.

2. The method of claim 1, wherein, The method further comprises the following steps: Whether the silence period is expired is judged, and if the judgment result is that the silence period is expired; Whether the fault of the corresponding alarm Internet of Things device is recovered is judged, and if the judgment result is that the fault of the corresponding alarm Internet of Things device is not recovered; Steps 1 to 6 are repeated until the fault of the corresponding Internet of Things device is recovered. The feature data comprises an Internet of Things device ID, an alarm type, a timestamp, a current alarm instantaneous frequency and system CPU / memory load, and the historical behavior mode comprises an alarm frequency, an alarm period and an alarm duration.

3. The method of claim 1, wherein, The method further comprises the following steps: Whether the corresponding Internet of Things device and the corresponding alarm type are in the silence state is checked; If the checking result is that the corresponding Internet of Things device and the corresponding alarm type are in the silence state, the corresponding alarm is discarded or aggregated.

4. The method of claim 1, wherein, The method further comprises the following steps: The effect of the suppression strategy is continuously monitored to generate feedback data; A feedback learning engine performs online optimization and adjustment on parameters of the pre-trained silence period prediction model based on the feedback data by using a reinforcement learning algorithm; The feedback data at least comprises an actual suppressed alarm number and whether an important alarm is mistakenly suppressed.

5. The method of claim 1, wherein, The real-time extraction of the feature data of the alarm and the query of the historical behavior mode of the corresponding Internet of Things device and the corresponding alarm from the historical database are realized by a behavior analysis engine in the Internet of Things cloud platform.

6. The method of claim 1, wherein, The reception of the feature data and the historical behavior mode, the input of the feature data and the historical behavior mode into the pre-trained silence period prediction model to output the dynamic silence period suggestion value of the corresponding alarm and the checking of whether the corresponding Internet of Things device and the corresponding alarm type are in the silence state are realized by a dynamic decision engine in the Internet of Things cloud platform.

7. The method of claim 1, wherein, The training steps of the pre-trained silence period prediction model comprise the following steps: Historical alarm data and associated system load indicators are collected; An instantaneous alarm frequency and a historical behavior mode are extracted as feature vectors, and an optimal silence period label is marked; A supervised learning or reinforcement learning model is trained to obtain the pre-trained silence period prediction model by taking the feature vectors as inputs and the optimal silence period as outputs.

8. An Internet of Things alarm dynamic self-adaptive suppression device, characterized in that, The device is applied to a system composed of an Internet of Things device cluster and an Internet of Things cloud platform, and the device comprises: A receiving module for receiving an alarm generated by the Internet of Things device cluster by the Internet of Things cloud platform; An extracting module for extracting feature data of the alarm in real time and querying a historical behavior mode of a corresponding Internet of Things device and a corresponding alarm from a historical database; An output module for receiving the feature data and the historical behavior mode, inputting the feature data and the historical behavior mode into a pre-trained silence period prediction model, and outputting a dynamic silence period suggestion value of the corresponding alarm, and checking whether the corresponding Internet of Things device and the corresponding alarm type are in a silence state; A setting module for setting a silence period with a length of the dynamic silence period suggestion value to suppress the corresponding alarm if the checking result is that the corresponding Internet of Things device and the corresponding alarm type are not in the silence state, and reporting the alarm to an operation and maintenance personnel; A judging module for judging whether the silence period is expired or not, and ending the suppression if the judging result is that the silence period is expired; An ending suppression module for judging whether a fault of the corresponding alarm Internet of Things device is recovered or not, and ending the suppression if the judging result is that the fault of the corresponding alarm Internet of Things device is recovered.

9. An electronic device, comprising: comprise: a processor and a memory; the processor is configured to execute the method according to any one of claims 1 to 7 by calling programs or instructions stored in the memory.

10. A computer-readable storage medium, characterized in that, the computer readable storage medium stores programs or instructions, which make the computer execute the method according to any one of claims 1 to 7.

Images